Conventional video search systems (e.g., video search engines, video hosting services, etc.) allow a user to search for videos responsive to a search query based on textual metadata associated with each video (e.g., titles, tags, descriptions, etc.). For example, if a user inputs a search query “Laplace,” a conventional video search system returns a list of links to videos with the word “Laplace” contained in their metadata, such as a video entitled “Laplace.”
However, such a conventional video search system may not provide a user with a satisfactory search experience for several reasons. For example, a conventional search system may return irrelevant search results by searching for videos using textual metadata that is not representative of the content of the video. In some instances, by inputting “Laplace” as a search query, a user may intend to search for video content in which the word “Laplace” is spoken, such as a video clip including a lecture on the Laplace transform. In such an example, search results generated based on textual metadata of videos, such as videos entitled “Laplace,” may be irrelevant to the search query inputted by the user.
As another example, a conventional video search system may return the same results responsive to a particular search query when a user resubmits the search query. That is, the user may have to continuously modify the search query and resubmit the modified search query to the video search system to view different search results. This can be a time consuming and frustrating procedure for the user, especially when the user is unaware of the search terms that may lead to the video content that the user is looking for.
Accordingly, it is desirable to provide new mechanisms for searching for video content.